{
  "id": 5018,
  "url": "https://nnemf.org/research/5018-statistical-analysis-of-personal-radiofrequency-electromagnetic-field-measurements",
  "title": "Statistical analysis of personal radiofrequency electromagnetic field measurements with nondetects.",
  "authors": [
    "Martin Röösli",
    "Patrizia Frei",
    "Evelyn Mohler",
    "Charlotte Braun-Fahrländer",
    "Alfred Bürgi",
    "Jürg Fröhlich",
    "Georg Neubauer",
    "Gaston Theis",
    "Matthias Egger"
  ],
  "journal": "Bioelectromagnetics",
  "year": "2008",
  "doi": "10.1002/bem.20417",
  "pmid": "18421711",
  "abstract": "Exposimeters are increasingly applied in bioelectromagnetic research to determine personal radiofrequency electromagnetic field (RF-EMF) exposure. The main advantages of exposimeter measurements are their convenient handling for study participants and the large amount of personal exposure data, which can be obtained for several RF-EMF sources. However, the large proportion of measurements below the detection limit is a challenge for data analysis. With the robust ROS (regression on order statistics) method, summary statistics can be calculated by fitting an assumed distribution to the observed data. We used a preliminary sample of 109 weekly exposimeter measurements from the QUALIFEX study to compare summary statistics computed by robust ROS with a naïve approach, where values below the detection limit were replaced by the value of the detection limit. For the total RF-EMF exposure, differences between the naïve approach and the robust ROS were moderate for the 90th percentile and the arithmetic mean. However, exposure contributions from minor RF-EMF sources were considerably overestimated with the naïve approach. This results in an underestimation of the exposure range in the population, which may bias the evaluation of potential exposure-response associations. We conclude from our analyses that summary statistics of exposimeter data calculated by robust ROS are more reliable and more informative than estimates based on a naïve approach. Nevertheless, estimates of source-specific medians or even lower percentiles depend on the assumed data distribution and should be considered with caution.",
  "effect": "unclear",
  "evidence": "moderate",
  "extraction": {
    "studyType": "exposure_assessment",
    "studyTypeLabel": "exposure assessment",
    "effectDirection": "unclear",
    "evidenceLabel": "Moderate",
    "population": "Participants in the QUALIFEX study (preliminary sample)",
    "sampleSize": "109",
    "exposure": "RF personal exposimeter (multiple RF-EMF sources) · weekly measurements",
    "confidencePct": "78",
    "peerReviewed": "yes",
    "mainFindings": "Using 109 weekly exposimeter measurements, differences between a naïve substitution approach and robust regression on order statistics (ROS) were moderate for total RF-EMF exposure (90th percentile and arithmetic mean). For minor RF-EMF sources, the naïve approach considerably overestimated exposure contributions, leading to an underestimation of the exposure range and potential bias in exposure-response evaluations; robust ROS was concluded to be more reliable, though source-specific medians/lower percentiles depend on distributional assumptions.",
    "outcomes": [
      "Comparison of summary statistics methods for RF-EMF exposimeter data with nondetects (robust ROS vs substitution at detection limit)",
      "Impact of nondetect handling on estimated total exposure and source-specific contributions"
    ],
    "limitations": [
      "Preliminary sample from the QUALIFEX study",
      "Robust ROS requires assuming a data distribution",
      "source-specific medians and lower percentiles depend on this assumption and should be interpreted cautiously"
    ],
    "keywords": [
      "RF-EMF",
      "exposimeter",
      "personal exposure",
      "nondetects",
      "detection limit",
      "regression on order statistics",
      "QUALIFEX",
      "summary statistics"
    ],
    "hubs": [],
    "raw": "{\n  \"study_type\": \"exposure_assessment\",\n  \"exposure\": {\n    \"band\": \"RF\",\n    \"source\": \"personal exposimeter (multiple RF-EMF sources)\",\n    \"frequency_mhz\": null,\n    \"sar_wkg\": null,\n    \"duration\": \"weekly measurements\"\n  },\n  \"population\": \"Participants in the QUALIFEX study (preliminary sample)\",\n  \"sample_size\": 109,\n  \"outcomes\": [\n    \"Comparison of summary statistics methods for RF-EMF exposimeter data with nondetects (robust ROS vs substitution at detection limit)\",\n    \"Impact of nondetect handling on estimated total exposure and source-specific contributions\"\n  ],\n  \"main_findings\": \"Using 109 weekly exposimeter measurements, differences between a naïve substitution approach and robust regression on order statistics (ROS) were moderate for total RF-EMF exposure (90th percentile and arithmetic mean). For minor RF-EMF sources, the naïve approach considerably overestimated exposure contributions, leading to an underestimation of the exposure range and potential bias in exposure-response evaluations; robust ROS was concluded to be more reliable, though source-specific medians/lower percentiles depend on distributional assumptions.\",\n  \"effect_direction\": \"unclear\",\n  \"limitations\": [\n    \"Preliminary sample from the QUALIFEX study\",\n    \"Robust ROS requires assuming a data distribution; source-specific medians and lower percentiles depend on this assumption and should be interpreted cautiously\"\n  ],\n  \"evidence_strength\": \"moderate\",\n  \"confidence\": 0.78,\n  \"peer_reviewed_likely\": \"yes\",\n  \"keywords\": [\n    \"RF-EMF\",\n    \"exposimeter\",\n    \"personal exposure\",\n    \"nondetects\",\n    \"detection limit\",\n    \"regression on order statistics\",\n    \"QUALIFEX\",\n    \"summary statistics\"\n  ],\n  \"suggested_hubs\": []\n}"
  },
  "synthesis": null,
  "synthesisStatus": "pending",
  "originalHref": "https://pubmed.ncbi.nlm.nih.gov/18421711/",
  "sourceUrl": "https://rfsafe.org/mel/paper.php?id=5018",
  "updated": "2026-02-14 08:11:46"
}